Graham-Dodd-Shiller 10 Year PE Ratio Stock Screen

Something I read last week got me interested in Professor Robert Shiller’s use of the 10 year price earnings ratio (PE) to value the stock market. It got me to thinking about whether this PE10 ratio could also be useful for valuing stocks.

Shiller first wrote about the use of long-term P/E ratios to value the market in his 1996 paper, “Price–Earnings Ratios as Forecasters of Returns: The Stock Market Outlook in 1996.” In that paper he examined the ratio of market price to the 30-year average of lagged earnings adjusted for inflation. In more recent work, Irrational Exuberance, Shiller used the inflation adjusted ten year average price earnings ratio, also referred to as P/E10 or Cyclically Adjusted Price Earnings Ratio (CAPE), to assess S&P 500 price levels relative to value.

Benjamin Graham and David Dodd described a similar ratio in their classic book, Security Analysis. They recommended taking price divided by the ten year average earnings to address fluctuations in earnings due to the business cycle.

I decided to try a similar PE10 ratio to rank stocks for further consideration. Specifically, I used an average inflation rate of 2.5% over the past ten years to adjust past earnings per share (EPS) to today’s dollars.  I took the average of the past ten years of inflation adjusted earnings and compared it to yesterday’s closing stock price for all publicly traded stocks. I then eliminated ADRs and stocks in the miscellaneous financial services sector. I also required that companies have a current ratio higher than 1.5 to avoid financially stressed companies. I also restricted the screen to companies with market caps greater than $500 million, PE10 greater than zero, and current PE less than 12. The 20 stocks with the lowest PE10 currently are listed in the table below.

Name

Ticker

PE10

CurRatio

PE

PB

ROI

The E.W. Scripps Company

SSP

5.16

2.32

4.28

0.92

2.93

Exide Technologies

XIDE

6.63

1.74

10.91

1.74

4.45

Constellation Energy Group

CEG

6.95

1.96

1.75

0.76

21.68

EnCana Corporation (USA)

ECA

7.68

1.69

9.61

1.18

5.31

SkyWest, Inc.

SKYW

8.95

3.02

11.35

0.62

2.06

Universal Corporation

UVV

10.59

2.27

8.17

1.06

8.98

RadioShack Corporation

RSH

11.18

2.02

10.72

2.38

14.83

Lexmark International, Inc.

LXK

11.57

1.78

9.35

2.23

14.03

P.H. Glatfelter Company

GLT

12.12

2.31

6.85

1.08

8.41

NRG Energy, Inc.

NRG

12.9

1.9

9.63

0.59

2.4

Computer Sciences Corp

CSC

13.38

2.16

8.87

1

6.53

Chevron Corporation

CVX

13.52

1.66

9.76

1.61

11.82

Exelon Corporation

EXC

13.74

1.7

10.01

1.88

5.69

Noble Corporation

NE

13.76

1.68

8.06

1.29

12.54

Eli Lilly & Co.

LLY

14.37

2.33

7.81

3.17

21.46

Garmin Ltd.

GRMN

14.74

4.16

7.77

1.93

24.78

Bunge Limited

BG

14.8

1.74

4.8

0.81

13.45

Archer Daniels Midland Co

ADM

14.91

1.81

10.53

1.21

7.5

Entergy Corporation

ETR

14.98

1.73

10.28

1.5

3.96

ENSCO PLC

ESV

15.1

3.19

10.51

1.19

9.41

A quick backtest of this screen from 2002 to present with a one year annual rebalancing period showed that $100 invested evenly in the stocks selected by the Graham-Dodd-Shiller PE10 Screen grew to $344 today compared to just $104 for the S&P 500 over the same period. This long-term value screen shows some real potential.

What do you think of the criteria used in this screen? I’d also be interested in hearing about your thoughts on the stocks currently selected by this screen.

Disclaimer: I do not currently have any position in the stocks discussed in this article.

11 thoughts on “Graham-Dodd-Shiller 10 Year PE Ratio Stock Screen

  • November 24, 2010 at 4:36 pm
    Permalink

    Very nice idea!

    Curious why you screened for stocks that had a current P/E below 12 though? Earnings could be depressed currently due to the recession, resulting in a high-looking current P/E.

    I’d love to see the results of this screen without that constraint, and without the minimum $500M constraint.

    Great idea, keep it up!

  • November 24, 2010 at 5:04 pm
    Permalink

    Agreed, great idea. I have to jump on the <$500m bandwagon. Most my investments these days are in the microcap stocks so would love to see the results from the tiny stocks as well.

  • November 24, 2010 at 5:39 pm
    Permalink

    Saj and Asues,

    Thanks for the feedback. If I look at stocks under $500 million in market cap, $100 invested on Jan 1, 2002 grows to $526 on Nov. 20, 2010! I’ll post a list of those stocks in another post shortly.

    I was actually looking for a mid to large cap value screen when I started exploring PE10. Many small caps don’t have 10 years of earnings, so I thought PE10 wouldn’t work well for them. Looks like I was wrong about that.

    Saj –
    As for restricting the screen to stocks with current PE (trailing twelve months) less than 12, my reasoning was that I wanted a list of stocks with strong current earnings yield. I just ran the screen backtest without the current PE<12 and $100 only grew to $121 from 2002 to November 20, 2010. It looks like capping current PE is important as I first guessed.

  • November 24, 2010 at 8:44 pm
    Permalink

    Hi George – interesting article with exciting results! I am wondering what software/website you are using to run those screens? I usually feel quite limited by the screeners out there and would love something with a bit more extensibility.

  • November 25, 2010 at 1:30 am
    Permalink

    what service did you use to perform the screen and backtest?

  • January 9, 2011 at 6:18 pm
    Permalink

    Are your backtests survivorship-bias free? That is, does the backtest find companies that might have passed that screen in the past but have gone bankrupt since and no longer have a symbol?

  • January 10, 2011 at 6:39 pm
    Permalink

    You should find the stocks that fit your criteria from the years, say, 1990-2000, and see how these stocks performed in the years since. The way you did it, you practically picked the winners based on your criteria. You need to see how past data predicts future data, not how past data predicts performance over the same period.

  • February 8, 2011 at 8:41 am
    Permalink

    Jeff – The backtests are survivorship-bias free. Companies that have gone bankrupt in the past are still included in the results.

    Scott – I wish I could run the screen from 1990 to 2000. However, I don’t have access to data that goes that far back. If anyone has access to that historical data please contact me and I’d love to do the backtest.

  • November 10, 2011 at 1:33 am
    Permalink

    Nice work —
    How abt using ur work on estimating the expected returns frm the stock as per formaula

    Investment Return (%) = Dividend Yield (%) + Business Growth(%) + (Mean_Shiller_PE/Current_Shiller_PE)(1/T)-1
    pl see this guru focus site fr further info and revert http://www.gurufocus.com/shiller-PE.php

  • January 17, 2012 at 7:37 pm
    Permalink

    George,
    I’ll start by saying great work and I appreciate all the effort you have put into researching the stock market. I recently read Phil Town’s “Rule #1 Investor” and have been inspired to get started in the stock market. I recently found some of your posts online, where you had run some calcs to find companies that fit the Rule 1 requirements. The posts were dated back in 2006. Was just wondering if you had any luck with any of those companies and also if you have put together a more recent list of Rule 1 companies.
    Thanks again,
    Mike

Leave a Reply

Your email address will not be published. Required fields are marked *